The literature on religious identification is filled with hyperbolic statements based on linear theories of social change. Reality is a lot messier, and driven more by fertility, immigration, migration, and socialization factors than anything approximating freely chosen religious commitments. Most people don’t change, and when they do, they don’t change much. Methodists become Baptists and vise versa. Presbyterians become Episcopalians or UCC. But you don’t see Pentecostals becoming UCC or vice versa. People prefer religious stuff that is familiar to their own experiences, and rewarded by the web of social connections. What is really cool about the new 2012 GSS is that we now have four full decades of high quality, comparable data. If we squish those out by the truly relevant categories of identification—which requires separating Episcopalians from other Liberals, and Lutherans from other Moderates, and Baptists from other Sectarians, and looking at Mormons and Unitarians as unique cases, you see a quite different perspective on the trajectory of American religious identification. First. while Catholics held a solid modal position with 26% of identifiers in 1972, they decrease their share to 23% in 2012—and Catholics are now only slightly more numerous than than people who reject religious affiliation —20% of Americans—and the non-religious now outnumber Baptists by a substantial margin. Baptists were a solid second place in 1972, with 20% of identifiers–but now trail nonidentification with only 14% of respondents in 2012. Baptists still manage third place, but with “no specific identification” Protestants hot on their heels at 12%—and those people are an ecclectic mix of disaffected conservatives, conservatives who are too stupid to figure out the denominational affiliation of their “nondenominational” church, and people who identify as Christian, but don’t give a rat’s ass much about religion. Claims that non-denominationals are all conservatives are bullshit–their rate of subscription to biblical fundamentalism is lower than what is found for moderate Protestants (42% versus 45%) and much lower than Baptists and sectarians (59% of Baptists and 64% of other sectarians are fundies). Becoming a non-denominational is more liberal than becoming a Methodist.
Notably, the “growth” among sectarian Protestants is limited–they took 7% in 1972 and 9% in 2012–and for “conservative churches” these moderate gains among the more fanatical sectarian groups are more than offset by the 6% loss found among Baptists. You can’t say conservative churches are growing by ignoring massive defection among Baptists. Mormons are not taking over the planet, and while the numbers are small here, the GSS found fewer Mormons in 1972 than in 2012. Decade groupings show minimal gains for Mormons between the 1970s and the 2000s—but for all of their attention, even the decade groupings find only three times as many Mormons as Unitarians—and nobody is saying that the Unitarians are taking over the world. Many liberal Protestant denominations declined, but the losses were much less substantial among Episcopalians, as I have shown in prior research. Lutherans and Moderate Protestants are also big losers.
The main reason for these shifts in distribution are fertility differences across generations, combined with differential losses to the big gainers—protestants with no denomination and people with no religious identification. I’ll be teasing out much more of this in my forthcoming book. But, the short story is that conservatives are not taking over the world—and a shitload of Americans have rejected religious identification. And, while religious operatives like to claim that people “believe but don’t belong”, rates of non-belief mirror non-identification. Over 20% of Americans reject any divine inspiration for the “bible” and over 20% of Americans are nontheists—being either atheist, agnostic, or believing in a higher power, but not a god.